📅 PUBMED DAILY TRIAGE — 2026-06-05

Window: 2026-06-04T09:00:00Z → 2026-06-05T09:00:00Z (CRDT) Topics scanned: 10 | Articles reviewed: ~57 candidates | Classified: 26 High priority: 4 | Standard: 18 | Low/archive: 4 | Unsolicited finds: 1 Run status: SUCCESS

🔴🟠 HIGH PRIORITY (8+) — Best Finds by Topic

Hematologic Malignancies + CBC/ML Diagnostics (T1+T2)

TR-2026-06-05-001 · PMID 42242261 · FIBOM-AI: Prediction of bone marrow fibrosis from CBC in myeloproliferative neoplasms · 9/10 🟢

Lancet Haematol multicentre ML study (n=2488 retrospective, n=493 prospective, 18 centres) demonstrates XGBoost model predicts BM grade 2–3 fibrosis from CBC + age with AUC 0.92 on external validation and 98.6% rule-out sensitivity prospectively — a direct, low-cost alternative or triage tool to invasive bone marrow biopsy in MPN.

Novel Therapeutics: Immunotherapy (T6)

TR-2026-06-05-002 · PMID 42243329 · Intratumoral bicarbonate as PD-1 blockade adjuvant in HCC · 8/10 🟠

Prospective clinical study (ChiCTR2100053537, n=30) of Tislelizumab + intratumoral 5% NaHCO3 in advanced HCC: ORR 93.3% (CR 53.3%, PR 40%), median PFS 31 months, OS not reached — extraordinary response rates via novel cGAS-STING/ICD mechanism. Requires large RCT confirmation.

Early Cancer Detection / Liquid Biopsy (T3)

TR-2026-06-05-003 · PMID 42243561 · Post-neoadjuvant ctDNA MRD in TNBC: systematic review and meta-analysis · 8/10 🔴

Meta-analysis of 22 studies (4 contribute primary estimate) finds ctDNA MRD positivity after neoadjuvant therapy associated with pooled HR 4.63 (95% CI 3.07–6.98, I²=0%) for recurrence-related outcomes in TNBC with residual disease — supports ctDNA as guidable MRD marker for escalation decisions.

AI/ML in Clinical Diagnostics (T4)

TR-2026-06-05-004 · PMID 42243600 · BrcaDetect: interpretable ultrasound AI for early breast cancer · 7/10 🔴 (HIGH by flag)

BrcaDetect integrates US images, BI-RADS, and demographics across 5 hospitals (n=3048); AUC 0.989 in training, 0.826 external; radiologist accuracy improved 6% (0.977 vs 0.919, p<0.001) with AI assistance; Grad-CAM + Shapley interpretability differentiates this from prior breast US AI tools.

🟢⚪ STANDARD (5–7) — Top by Topic

Hematologic Malignancies (T1)

• TR-013 · PMID 42243281 · CircRNA PTPN22/CARM1 axis in ALK+ ALCL chemoresistance · 5/10 · Drug Development
• TR-014 · PMID 42243643 · Pediatric AL epidemiology in Afro-descendant Caribbean population (n=107, 5y-OS 90.9%) · 5/10 · Public Health [UNDERSERVED]
• TR-023 · PMID 42243411 · SEM symptom burden in 40,766 cancer survivors incl. lymphoma · 5/10 · Public Health
• TR-015 · PMID 42243010 · CV complications in MPN — hematologists' review · 5/10 · Diagnostics

Early Cancer Detection / Liquid Biopsy (T3)

• TR-005 · PMID 42243127 · Methylation-based ctDNA monitoring in metastatic breast cancer during CDK4/6i therapy (Nat Commun, n=57, HR 0.17 for ctDNA clearance, 5.8 months lead time) · 7/10 · Diagnostics
• TR-006 · PMID 42243550 · SNP-NIPT prospective cytogenetic validation n=4466 (T18 sensitivity 100%, T21 80%, specificity 99.9%) · 7/10 · Diagnostics

AI/ML in Clinical Diagnostics (T4)

• TR-008 · PMID 42243228 · Weakly-supervised DL for cutaneous vasculitis WSI pathology AUC 98.39%, 2 centres · 6/10 · Diagnostics
• TR-020 · PMID 42243352 · DA-GAT-v2: ECG fairness AI, sex gap 15.4%→1.8%, AUROC 0.9762, PTB-XL 21,507 recordings · 5/10 · Diagnostics

Precision Oncology (T5)

• TR-018 · PMID 42243754 · TIM-3 as independent DFS predictor in ICC n=117 (HR 1.589) — potential immunotherapy target · 5/10 · Genomics/Precision Medicine
• TR-025 · PMID 42243732 · CA125/HE4/ROMA/CPH-I in advanced EOC: ROMA/CPH-I AUC ~0.80 for cytoreduction prediction · 5/10 · Diagnostics

Novel Therapeutics (T6)

• TR-009 · PMID 42242867 · ACTION-1 [225Ac]Ac-DOTATATE phase 1b/3 dosimetry (NCT05477576) — favourable tumor-to-normal ratio in GEP-NETs · 6/10 · Treatment Innovation
• TR-021 · PMID 42243532 · PLD1/PLD2 TME immunosuppression via CCL19/PD-L1 — melanoma preclinical · 5/10 · Drug Development
• TR-022 · PMID 42243682 · Lung microbiota-PD-1/PD-L1 correlations in 350 lung adenocarcinoma patients · 5/10 · Treatment Innovation

Cardiovascular-Metabolic (T7)

• TR-010 · PMID 42243606 · hiPSC-CM vs animal cardiac QT risk assessment (FDA pilot, concordance 0.71–0.82) · 6/10 · Drug Development [NEAR_TERM]
• TR-016 · PMID 42243742 · RAR-MASLD mortality NHANES n=6287 (HR 2.41 all-cause, HR 3.39 CV) · 5/10 · Diagnostics
• TR-017 · PMID 42243686 · ΔmtDNA as STEMI prognostic biomarker n=236 (HR 1.239/SD) · 5/10 · Diagnostics
• TR-024 · PMID 42243704 · 71.2% uncontrolled SBP in treated elderly Sri Lanka n=1339 [UNDERSERVED] · 5/10 · Public Health
• TR-026 · PMID 42243551 · ALT low sensitivity (19.2%) for UGAP-steatosis in obese children n=91 · 4/10 · Diagnostics

Rare Diseases (T9)

• TR-007 · PMID 42242766 · CCHS European Consortium survival n=240: HSCR → HR 6.8 mortality (25y-OS 26% vs 89%) · 7/10 · Public Health [UNDERSERVED]
• TR-012 · PMID 42239344 · GREGoR rare disease multi-omic data model n=12,292 participants (bioRxiv) · 6/10 · Genomics/Precision Medicine
• TR-019 · PMID 42242566 · FOP NIV in 3 children: long-term tolerance + reduced exacerbations in ultra-rare disease · 5/10 · Treatment Innovation [UNDERSERVED]

💡 UNSOLICITED FINDS

TR-011 · PMID 42243831 · Prolonged SAMe exposure associated with poorer breast cancer treatment response (n=1013, PSM-adjusted) · 6/10 · Treatment Innovation [PROMISING_PRELIMINARY]

Sentinel-detected: SAMe hepatoprotection (≥14 days) during adjuvant chemotherapy independently associated with shorter OS/DFS after propensity matching; intratumoral m6A/METTL3/METTL14 elevated in poor-responders. Practice implication for SAMe use in Asia/China if confirmed prospectively.

📊 TODAY'S BREAKDOWN

By research category:

• Diagnostics: 12
• Treatment Innovation: 7
• Drug Development: 3
• Public Health: 5
• Genomics/Precision Medicine: 3
• Early Detection (as primary): 1

By evidence maturity:

• Exploratory: 20
• Validated: 5
• Potentially Practice-Changing: 0
• Preprint (capped): 1

By watchlist topic (reviewed / kept / highest score):

• T1 Hematologic malignancies: 10 / 5 / 9 (FIBOM-AI)
• T2 CBC/ML: 1 / 1 / 9 (FIBOM-AI, crossover)
• T3 Liquid biopsy/ctDNA: 6 / 3 / 8 (ctDNA MRD TNBC meta-analysis)
• T4 AI/ML diagnostics: 5 / 3 / 7 (BrcaDetect)
• T5 Precision oncology: 8 / 2 / 5 (TIM-3 ICC)
• T6 Novel therapeutics: 8 / 4 / 8 (Bicarbonate+PD-1 HCC)
• T7 Cardiovascular-metabolic: 5 / 5 / 6 (hiPSC-CM FDA study)
• T8 Aging/longevity: 0 / 0 / 0 (no results — zero-result run)
• T9 Rare diseases: 8 / 3 / 7 (CCHS survival)
• Sentinel: 5 / 1 / 6 (SAMe breast cancer)

By source type: peer_reviewed: 25 | preprint: 1

Topics with zero 8+ articles this run:

• T4 (best was BrcaDetect 7/10, HIGH by flag only)
• T5 Precision oncology (best 5/10)
• T7 Cardiovascular-metabolic (best 6/10)
• T8 Aging/longevity (zero results)
• T9 Rare diseases (best 7/10)
• Sentinel (best 6/10)

🔧 RUN NOTES

• T8 Aging — zero results (second consecutive run): Updated query with telomere/inflammaging/epigenome additions still yields zero in the CRDT window with humans[MeSH]. Hypothesis: humans[MeSH] filter is suppressing newly-indexed records not yet fully MeSH-tagged. Recommendation: Test query without humans[MeSH] in next run as diagnostic; if false-positive rate returns, add NOT (animals[MeSH] NOT humans[MeSH]) as alternative filter.

• T3 truncation: 4 ctDNA articles (PMIDs 42242004, 42241805, 42241155, 42240938) from the 26-hit result set were not retrieved due to batch size constraints. These may include additional significant findings from the active liquid biopsy literature.

• Sentinel volume extremely high (1279): Only top 5 reviewed. Sentinel query is too broad to be meaningfully curated — consider adding AND ("Randomized Controlled Trial"[PT] OR "Clinical Trial, Phase III"[PT] OR "Meta-Analysis"[PT]) to reduce noise and focus on high-design-quality findings.

• T7 GLP-1/SGLT2 RCT gap: No primary GLP-1/SGLT2 RCT data in top 5 reviewed this window. Cardiometabolic coverage dominated by observational/biomarker studies. Adding AND ("randomized controlled trial"[PT] OR "clinical trial"[PT]) optionally may enrich trial data when query volume is high.

• T2 CBC/ML — FIBOM-AI is outstanding: The single T2 result today (PMID 42242261) was the top finding of the entire run. T2 query now reliably avoiding false positives since the 2026-06-04 refinement. Volume remains low (1 article today) — this is expected for a specific intersection.

• T5 Precision Oncology noise: High volume (232 raw) produced only 2 STANDARD articles from top 8 reviewed. The T5 query is too broad — biomarker and targeted therapy alone without cancer type restriction yield too much noise. Recommend restricting to ("targeted therapy"[tiab] AND (cancer[tiab] OR tumor[tiab] OR carcinoma[tiab])) OR "precision oncology"[tiab] OR "actionable mutation"[tiab] OR "tumor mutational burden"[tiab] in next run.

• Window expansion NOT triggered: 26 relevant records retrieved (≥5 threshold met); no expansion to 48h required.

• Deduplication check: No overlap with 2026-06-04 emitted fingerprints confirmed.